A molecular component to epidemiological studies can help to identify homogeneous patient subgroups, provide dosimeters of exposure, and generate new knowledge of the biological plausibility of etiologic hypotheses. Wilms' tumor is a particularly useful model for studying the molecular epidemiology of cancer in humans. The postulated existence of at least 3 candidate 'Wilms' tumor genes' typifies the genetic heterogeneity of the common human cancers. One Wilms' tumor gene, the WT1 gene, was recently isolated. It displays the cardinal features of a tumor suppressor gene, one of the 2 major classes of genes in cancer development. Studies by us and others have detected intragenic germ line WTl mutations in a few Wilms' tumor patients who have bilateral neoplasms or associated urogenital anomalies. To pursue these observations, we propose herein to conduct an analytical epidemiology study of germ line WT1 mutations in a larger series patients to identify the molecular basis for several major epidemiological features of Wilms' tumor. We postulate that the epidemiological findings of familial aggregation of the tumor, bilaterality, and association with certain urogenital anomalies are due in part to WT1 gene mutations. Furthermore, the specific epidemiologic characteristics in a patient are determined in part by location and size of the mutation in WTI and their effects on the WT1 protein product. The study has a case-control design (cases have either 1- familial or bilateral Wilms' tumor, or 2-Wilms' tumor and associated urogenital anomalies and controls are Wilms' tumor patients with no associated condition). The blood specimen collection from these rare cases will be through the National Wilms' Tumor Study (NWTS), which has accrued nearly 5,000 children in the clinical treatment trials (approval received). Detection of germ line mutations in the WTI gene will be achieved by Southern blotting, single-strand conformational polymorphism analysis of the 10 WT1 exons, ana gene sequencing. Results will become part of a Registry of Germ Line WTI Mutations, including a DNA bank that will be useful in the search for the 2 other putative Wilms' tumor genes.